EP0413123A1

EP0413123A1 - Method of transmitting data over an optical star network, and station and optical star network operating according to the same

Info

Publication number: EP0413123A1
Application number: EP90112834A
Authority: EP
Inventors: Gianfranco De Cristofaris; Franco D'ignazio
Original assignee: Alcatel NV
Current assignee: Alcatel Lucent NV
Priority date: 1989-07-06
Filing date: 1990-07-05
Publication date: 1991-02-20
Anticipated expiration: 2010-07-05
Also published as: ATE117488T1; JP2758068B2; ES2070217T3; FI102339B1; IT1230301B; FI903400A0; JPH0345044A; DE69016126T2; FI102339B; GR3014921T3; EP0413123B1; US5172374A; IT8921112A0; DE69016126D1

Abstract

The present invention relates to a UCOL-type star network, comprising a central unit or passive star center and a plurality of depending stations intercommunicating via two-way optical fiber link channels, in which an access protocol generated by the stations linked to the star center shows a frame structure cyclically repeating every millisecond,and is shared in at least three parts where, in the first one,informations related to frame beginning are idividuated,in the second one informations related to priority queues status are individuated, which informations allow to determine which station has access to the network, and are inherent to priority classes allowing to manage the various kinds of synchronous and asynchronous traffic, as well as a third part in which there is the information content to be transmitted, the whole under a constant control on data flow synchronism,in order to avoid a mutual data overlapping and to allow the sending of various kinds of traffic on the same channel, upon initialization of each station with an own time information depending on its distance from the star center.

Description

The present invention relates to a UCOL-type star network for the optimal management of various kinds of traffic according to the preamble of claim 1.
As it is known,in the communications management in a network,where there are several transmitting stations,it is important to determine the access priorities of each station to the network itself,and then to the resources thereof,so as to avoid collision between the informative packets generated by the various stations.
Relatively to UCOL networks, such a problem is particularly prominent due to their particular characteristics concerning topology, technical features and types of offered services.
Indeed,a UCOL network is configured as a passive star where each station is connected with its star center via two dedicated optical fibres.One fibre serves for transmission from the station to the star center and the other one for reception.
Each optical fibre is assimilable to a multiple transmission channel.
The particular kind of optical fibre link using the coherent optics technology,allows that different channels be multiplexed through frequency division and used for several transmissions along the same fibre, thus optimizing the link resources.
As concerns the management of the access of each station to the network,the present access protocols do not fit with a UCOL network.Moreover,although the access protocol systems using the priority concept are already available,at the present state of the art they appear to be intrinsecally very complicated as to frame structure.For such reason they cannot satisfy the functional characteristics of a UCOL network and cannot optimize the resources therein con tained.
This is due to the fact that the trend of the ideal management of access disputes to network resources, during the transmission of the various data coming from the various stations, is not well defined in the prior art.
Therefore,the object of the present invention is to eliminate the drawbacks of the prior art and to provide a UCOL network able to meet all the mentioned requirements by using new access protocol features which allow to solve the access dispute to the resources of a type UCOL network and to optimize the network resources,as well as to be able to transmit simultaneously,along the same channel and through cell transmission, both synchronous and packet traffic.
This object is achieved by the features set forth in the claim 1,whereas further details and improvements are evidenced in the subclaims.
From an operative point of view, such a network can operate both with synchronous and asynchronous traffic, by using the same channel (reference is made to packet and synchronous traffic as an example).
Moreover, with such a network it is possible to avoid the collision between data flows, being able to provide for the errors due to information on the measurement of distance between the station and the star center,and it is possible to prepare the informations for transmission, knowing a transmission advance time which depends on the information propagation delay along the optical channel in addition to the measured distance between the station and the star-center.
Finally with such a network it is possible to classify, thanks to priority classes, the kind of traffic to be managed.
The invention will now be described in detail next herein, by way of an example, with reference to the attached drawings, wherein:

-in Fig.1 a UCOL network is schematically illustrated; and
-in Fig.2 there is illustrated the frame structure of the access protocol of the UCOL transmission network, according to the present invention.

First of all it is a good thing to state beforehand that generally a UCOL network is to be considered as a multichannel network,thereby the access protocol described hereinafter is referred only to one channel,but what has been stated can be implemented on all others.
The ideal situation for a UCOL network is the one in which there is a continuous data flow, controlled by the access protocol from the stations 7,8,9,10...N to the star-center 6 without these data packets colliding each other thus generating an unintelligibility of their contents.
The access protocol,according to the present invention,provides for the subdivision of each channel 1 into frames 2.
First of all, an optimal repetition time of each single frame,fixed in one millisecond,i.e. each frame repeats at each millisecond in the transmission channel, is established. The frame 2, according such an access protocol, is then shared out,in this istance,into three parts. More particularly, into a first part 3 individuating basically an information related to frame beginning; into a part 4 characterizing essentially an information about the status of data access queues from the various stations; as well as into a part 5 related to data informative contents from the various stations.
Each of these parts is, in turn, subdivided into sections having a well defined quantity of bytes.
The part 3 of the beginning of frame,specifically, comprises 48 bytes and is subdivided in two sections 3a and 3b. The section 3a contains the information related to frame beginning.The section 3 contains a field called GUARD TIME related to a time approximately evaluated as 50 nanoseconds which provides for measurement errors related to a time information concerning the measurement of distance between a station and the star center 6,of which an exhaustive explanation will be given hereinafter,in order to avoid colli sion among the data blocks.
In part 4 of frame 2,related to queues status, are a first section 4a,relative to the synchronization of data flow from the stations, which acts for discriminating several blocks transmitting in synchronism; a section 4b, called MGT (MANAGEMENT), which is an additional section whose information acts for providing new communication channels for distributed networks upon new requests from management of the network itself.
In part 4 are then five more sections 4c, 4d, 4e, 4f, 4g, which, indicated in figure by P0 to P4, represent five priority classes related to access priority for data coming from the stations to the star center, used for selecting data access of a station with respect to the others, as a function of transmission priority, i.e. of wait time. All this to conform the synchronous traffic treatment modes to the packet traffic treatment mode, but considering that the synchronous traffic has a higher priority with respect to packet traffic.All the above mentioned traffics are thus managed by informations related to corresponding priority classes.
Normally, the higher priority class is that one having the lower number,i.e. class 0 represents the highest priority. Morevoer, class 0 is exclusively reserved for synchronous traffic,while the other ones are reserved for further priorities. Each priority class has two bytes for quantizing the priority intrinsic value and which in any case represents, as priority class, a number proportional to the maximum time the information transmitted by a station, meant as data quantity, must wait before its transmission. Normally, this time decreases as the wait time increases and, when it reaches the maximum priority and is not transmitted, the packet is abandoned.
In part 4 according to access queues status in the frame, it is decided which of the packets must be transmitted with the subsequent frame.
Now ,since in the present case the maximun foreseen number of stations is better to be 64, in such part there are 64 subparts, each structured as part 4 just described from 4a to 4h.
Also in part 4,at the end of the five priority classes,there is a GUARD TIME field, in that case indicated by 4h,having the same purposes of the previous one, namely 3b.
Each of the 64 subparts of part 4, configured as hereinbefore described, contains totally 28 bytes.
For the synchronization,16 bytes are needed,while for section 4b,2 bytes are needed; on the contrary, for each priority class,2 bytes are necessary.
Finally,the part 5 related to informations is also structured in various sections: one,5a,for synchronism,as said in section 4a, another one containing an information characterizing both the source and the destination of the informative contents and the information beginning, and is indicated by 5b,and,in the end, there is the cell information,indicated by 5c, still closed by GUARD TIME field 5d. The whole part 5 can contain 200 subparts, each structured in the same manner as part 5, from 5a to 5d, just illustrated and provided with 88 bytes for each of 200 subparts.
Of these 88, 16 bytes are used for synchronization, 3 bytes for identifying the terminal station and for having a spare for eventual applications not foreseen and, of the remaining bytes, 64 are used for the real information, also in conformity with the rules nowadays in use, and five for providing a further beginning of real intrinsic informative part.
Obviously, several informations pertaining to more than one of such 200 subparts, can be grouped together so as to provide a single informative concept.
After each frame has been prepared and after each station is enabled to send its informations, the entire network will have its map of informative cells wherein each station will have priorities characterized by priority classes 4c to 4g.
All this occurs upon initialization of each station with an own time information depending upon its distance from the star center.
Indeeed,each station sends initially an own signal toward the star center,waiting its return. In this way it is possible for each station to get the initialization time information, which value corresponds with the distance of the station from the star center. In substance, said time information characterizes an advance time for the transmission of data informative contents, which time begins when allowed by the synchronism information and the priority class and lasts for a period corresponding exactly to propagation delay of the information along the link optical channel and depends on the effective measured distance between the station and the star center.
The bytes present in the various sections of the various parts 3,4,5 provide a data set which,travelling along channel 1, allow to avoid collisions among data coming from the various stations,optimize the channel sharing trough the one millisecond time, and allow to provide an information on the status of accesses through queues status 4.
Moreover, cell transmissions can also be sent, as just mentioned, both in case of synchronous traffic and packet traffic.
Such a protocol structure for network access can be easily implemented because it can be subdivided into further frame shares without any difficulty.Besides, the frame structure is particularly simple and does not imply any complication neither of bytes or further subparts burden.
The basic concepts of the network subject of the present invention can be summarized,on the one hand, in the initialization of stations according to their distance from the star center and,on the other hand, in the frame sharing into three parts 3,4,5, which provide together a plurality of characteristic informations,the access queues status, as well as the possibility of cell transmission, which make such type of protocol specific for UCOL networks. Indeed, such networks have the particular star topology, as well as particular technical aspects such as distance and station number parameters, and the aforementioned services such as, inter alia, the possibility of managing simultaneously, on the same channel, both the synchronous traffic and the packet traffic.
Relatively to the UCOL network described in the present invention, the number of usable optical channels has been fixed equal to 24.
It must be specified that the maximum number of stations linkable to the star center via optical channels should never exceed thirty kilometers, since such parameters should largely influence the network performances.
Obviously, other improvements and developments are also possible, all falling within the inventive sphere of the present invention.

Claims

1) A UCOL-type star network for an optimal management of various kinds of traffic, of the type comprising a central unit or passive star center and a plurality of depending stations connected via two-way optical fiber channels,said stations communicating trough an access protocol comprising a frame structure, said network being characterized in this, that the stations, once assumed an initialitation time information, which is characteristic of each station and function of its distance from the star center, are able to generate an own access protocol wherein each frame is presented in a sequential fashion and cyclically repeated in time with a duration fixed and structurally divided into several time parts in which it is possible to individuate informations concerning the beginning of the frame, the priority queues status to determine the access to the network of the stations data and the priority classes corresponding to various kinds of traffic, as well as the informative contents to be effectively transmitted by each station, once a synchcronism information of data flow, cyclically transmitted by one of the stations, has obtained the consent for transmitting the informative data contents.

2) A star network according to claim 1, characterized in that said time information individuates an advance time of the informative contents, present in data which begins when allowed by the synchronism information and by priority class and lasts for a period corresponding exactly to the propagation delay of the information along the link optical channel and depends on the effective distance measured between the station and the star center.

3) A star network according to claim 1, characterized in that the number of said priority classes is equal to the various kinds of traffic which can be managed by said network.

4) A star network according to claim 1, characterized in that the informative data contents is transmitted in cell modules.

5) A star network according to claim 1, characterized in that said time parts are at least three, in each frame, namely a first part in which there are identified a first section containing said information related to frame beginning, and another containing a further field,called G.T.(GUARD TIME) ,to provide for the errors related to the information about the measurement of the distance between station and star centers ; a second part in turn divided into a number of subparts equal to the number of stations manageable in said network, in each of which there are identified a section containing said information related to data flow synchronism,a series of sections containing said informations related to priority classes and to priority queues status, and a final section containing a further field of aforesaid G.T. type; as well as a third part in turn subdivided into a plurality of subparts, in each of which there are identified a section containing the information related to data flow synchronism, a section assigned to the intrinsic informative contents and a final section containing a further field of aforesaid G.T. type.

6) A star transmission network according to claim 1, characterized in that the fixed repetition time of the frame is of one millisecond.

7) A star transmission network according to claim 5, characterized in that in each subpart of the second part there is also a further section, called MGT (MANAGEMENT) containing an information representing an additional section,which provides new transmission channels upon new de mands in the network management.

8) A star transmission network according to claim 5, characterized in that in each subpart of third part there is also a further section containing the information related to the source of the informative contents transmitted with the cell module.

9) A star network according to claims 1 or 5, characterized in that the first part contains a total number of bytes equal to 48.

10) A star network according to claim 5, characterized in that each subpart of the second part has a number of bytes equal to 28.

11) A star network according to claim 5, characterized in that each subpart of the third part has a number of bytes equal to 88, 69 of which belonging to the informative module cell.

12) A star network according to what hereinbefore described and illustrated for the specified objects.